China's new open-source model accelerates AI hacking threat
Summary
A new Chinese open-source AI model called GLM-5.2 matches the capabilities of some of the top U.S. AI models but costs much less to run. Security experts are worried this lower cost and open access will make advanced hacking tools easier for criminals to use and customize.Key Facts
- GLM-5.2 is an open-source AI model released last week by Chinese company Z.ai.
- It performs similarly to leading U.S. AI models like GPT-5.5 and Claude Opus 4.8 in cybersecurity tasks.
- Some researchers claim GLM-5.2 may be illegally derived from these U.S. models, explaining its high performance.
- Open-weight models like GLM-5.2 can be freely downloaded, changed, and used without safety limits.
- Hackers are already sharing ways to jailbreak GLM-5.2 to bypass restrictions and use it for cyberattacks.
- GLM-5.2 allows hackers to create personalized attacks, including generating phishing emails and fraud scripts.
- Open-source models offer no central provider to detect or block malicious use, unlike commercial AI services.
- Chinese companies plan to release more powerful open-source AI models later this year.
Read the Full Article
This is a fact-based summary from The Actual News. Click below to read the complete story directly from the original source.